10 Gigabit Ethernet is an increasingly popular communication standard, with a nominal data rate of 10 Gbit/s. One form of 10 Gigabit Ethernet is IEEE 1OGBASE-T, used to provide 10 gigabit per second connections over unshielded or shielded twisted pair copper wires. Each 10GBASE-T transceiver channel typically includes a transmit path and a receive path. As signals are transmitted, echos or reflections may result and propagate back along the transmit path, forming an “echo channel.” Forward signal propagation occurs in what is often referred to as a “forward channel.”
In a full-duplex system, the echo channel often depends on components external to the device that couple the analog transmit path to the analog receive path. Although transmit path high-frequency distortions are generally filtered out through the cable forward channel, this may not be the case through the echo channel. Thus, there may be a detrimental impact on local receive performance without impacting remote receiver performance, especially for long cable lengths.
Besides the echo channel performance noted above, there are many other parameters in a 10GBASE-T transceiver that need to operate near an optimal operating point in order to robustly transmit and detect data to and from the link. The optimal operating point is usually determined through an initial training process where a far-end transmitter sends a known training signal and a local receiver optimizes its own parameters using the training signal and a training method. Some of the transceiver parameters that benefit from training, or calibration, include transmit phase, analog-to-digital conversion parameters (such as gain matching, and offset cancellation/matching), driver parameters, transmit gain, and termination impedance, among other things. The training methods typically find the optimal receiver parameters by minimizing some measure of error. The error is usually identified as the difference of the known transmit data and the received data decoded by the receiver.
While initial training methods often work well for their intended applications, several transceiver operating parameters are analog in nature, and may be susceptible to PVT variations over time. PVT variations often undesirably affect transceiver performance. Unfortunately, known calibration methods for high-speed ethernet transceivers do not identify and/or compensate for PVT effects at both the chip characterization stage and in the field.